Sources of Linked Data
http://www.gutenberg.org/wiki/Gutenberg:Feeds Project Gutenberg Feeds
Project Gutenberg's entire catalog is available as a single RDF file (around 200MB unpacked), or one file per book/entry.
The dbpedia project aims to automate the translation of wikipedia data into RDF. There is a SPARQL endpoint.
A commercial service for creating community-driven "Wikipedia for Data", Developed by the company Metaweb, the project and company were purchased privately by Google in 2010
CYC is a classic Artificial Intelligence project to encode all "common sense" knowledge to support Machine intelligent software projects. Now available for the Semantic Web.
The Flemish Theatre Institution offers data in RDF.